The present invention relates to an image pickup apparatus. More particularly, the present invention relates to a compact and low-cost image pickup apparatus that is capable of photographing high-quality images comparable to silver halide photographs and particularly capable of suppressing the image quality degradation due to color shift caused by the lateral chromatic aberration of a lens.
There are digital cameras designed for business use, e.g. printing, which are capable of obtaining images comparable in quality to photographs. These digital cameras have a large number of pixels, which exceeds 2 million pixels. However, the pixel pitch of the image pickup device is as large as about 7.5 to 10 micrometers or more. Accordingly, there is a limit to the number of devices obtained from one wafer. Therefore, the cost cannot be reduced in excess of a certain limit.
The most effective way of reducing the production cost of image pickup devices is to reduce the device size so that an increased number of devices are obtained from one wafer. However, it is necessary in order to reduce the device size to reduce the number of pixels or the pixel pitch.
Many of image pickup apparatus that have been commercially manufactured as relatively low-cost apparatus use image pickup devices of ⅓ to ⅔ inch size and with about 1 to 2.5 million pixels. The pixel pitch of these image pickup devices is about 3 to 5 micrometers, which is markedly small in comparison to image pickup devices used in the above-described digital cameras for business use.
To obtain a high-quality image with an image pickup device having a large number of pixels, it is essential to use an image pickup device with a smaller pixel pitch.
Major factors in determining high image quality include color shift in addition to resolution. It is known that even if the image-forming performance for a single wavelength is high, when there are large chromatic aberrations, loss of color definition and color shift occur, causing the image quality to be degraded. In the case of an image pickup apparatus using a solid-state image pickup device, e.g. a digital camera, in particular, color shift caused by lateral chromatic aberration rather than axial chromatic aberration becomes a serious problem.
FIG. 12 shows the relationship between the input light intensity (analog image) and the output intensity signal from each pixel. As shown in the figure, in a device in which images formed in a certain fixed area are averaged or added together to form an image signal, such as a solid-state image pickup device, even when an image with gradient light intensity, the resulting image gives a xe2x80x9cstairstepxe2x80x9d appearance.
In silver halide photography, on the other hand, images are composed of dye units varying in dye size from several micrometers to several tens of micrometers. Thus, the image intensity approximately reproduces the gradient of the light intensity.
In comparison of the above two, there is only a small difference at a position where the gradient of the light intensity is small. However, there is a large difference at a position where the gradient of the light intensity within one pixel is large (e.g. the position of the second pixel from the left). That is, in the case of using an image pickup device, an area where the light intensity is low in the actual image (e.g. the right-hand end in the second pixel from the left) undesirably becomes an image having an average light intensity in the pixel.
This is not a serious problem in a monochrome image or an image in a range where the hues are similar to each other. That is, in such an image, the gradation of the image merely becomes stairstepped. However, in a case where only a specific color is displaced, as shown in FIG. 13, the color difference gives a xe2x80x9cstairstepxe2x80x9d appearance. For the sake of simplicity, only two colors, i.e. R (red) and B (blue), are considered in FIG. 13. In a case where the image of B is displaced with respect to the image of R as shown in part (a) of FIG. 13, the difference between the colors R and B is as shown in part (b) of FIG. 13. Thus, the color difference signal intensity R-B becomes stairstepped.
It is known that human beings are more sensitive to a color difference than to the absolute value of a color. Therefore, if a human being observes such an image, the color shift appears to be larger than the actual shift. Accordingly, it has been demanded that a lens used in an image pickup apparatus using an image pickup device should be highly corrected for lateral chromatic aberration that causes color shift.
Various studies have heretofore been conducted to determine a size of lateral chromatic aberration necessary for the above-described intensity ratio between two colors displaced relative to each other to appear to be no large color difference to the observer. As a result, it is known that a lateral chromatic aberration not larger than xc2xd of the pixel pitch allows the color shift of an image to be substantially ignorable.
However, if it is intended to construct an image pickup apparatus capable of providing high-quality images at reduced cost by using a small-sized image pickup device, the pixel pitch becomes 3 to 5 micrometers or less, as stated above. If such an image pickup device is corrected for lateral chromatic aberration on the basis of the conventional concept, the allowable lateral chromatic aberration is 1.5 to 2.5 micrometers or less. Accordingly, an extremely high level of chromatic aberration correction is required.
In the case of an image pickup apparatus using an image pickup device, filters, e.g. an optical low-pass filter and an infrared cutoff filter, are placed between an optical lens and the image pickup device. In the case of a multi-chip type apparatus, an optical path branching prism or the like is frequency placed therebetween. Accordingly, although the physical size is small, a long back focus is required from the optical point of view. Therefore, the common practice is to use a lens type that allows the back focus to be lengthened. Consequently, in many cases, the power distribution is asymmetrical with respect to the stop as viewed from the object side toward the image side.
Accordingly, it is difficult to correct aberrations depending on the field angle, such as distortion and lateral chromatic aberration. There has heretofore been a known method of correcting chromatic aberrations by using anomalous dispersion glass.
To correct lateral chromatic aberration to a value within the above-described range, it is necessary to further increase the number of lens elements using anomalous dispersion glass.
However, anomalous dispersion glass generally has a large coefficient of expansion or has such properties that it is soft and readily scratched. Therefore, a large number of man-hours are needed to process anomalous dispersion glass. Thus, using a large number of lens elements of anomalous dispersion glass causes the processing cost to increase unfavorably.
To realize a high level of chromatic aberration correction with a reduced number of lens elements of anomalous dispersion glass, it is necessary to weaken the power of each lens element by increasing the number of lens elements used or increasing the size of the lens system. This interferes greatly with the achievement of a reduction in size of the apparatus.
Meanwhile, an image obtained by using an image pickup device can be handled spatially as digital data.
Therefore, if the intensity of the image is digitized, it becomes easy to subject the image to electrical processing, e.g. image processing. Accordingly, a method wherein chromatic aberrations produced in a lens system are corrected by image processing has been proposed [Japanese Patent Application Unexamined Publication (KOKAI) No. 8-205181]. However, if image processing is carried out, it becomes necessary to perform processing for preparing data at image points where there is no data in the original image. Consequently, the image quality is unavoidably degraded. Moreover, as the processing becomes complicated, an unfavorably long time is needed until the data is recorded on a recording medium, e.g. a memory card, after the completion of photographing. In addition, because the characteristics of a lens are utilized, different processing operations are required for lenses having different characteristics. Accordingly, different processing operations corresponding to the number of lenses used are needed.
As has been stated above, to obtain images comparable to silver halide photographs, in which color shift is inconspicuous, the prior art causes various problems, e.g. an increase in- processing cost, an increase in size of the lens system, and an increase in image processing time after photographing, and cannot attain a compact and low-cost image pickup apparatus.
In view of the above-described problems associated with the prior art, an object of the present invention is to provide a compact and low-cost image pickup apparatus capable of photographing high-quality images comparable to silver halide photographs, particularly an image in which color shift is inconspicuous.
To attain the above-described object, the present invention provides an image pickup apparatus wherein an image of an object produced by an optical system is formed on an electronic image pickup device, thereby obtaining image information concerning the object. The electronic image pickup device and the optical system satisfy the following conditions:
1.5 less than P less than 7.0xe2x80x83xe2x80x83(1)
3.0 less than N less than 20xe2x80x83xe2x80x83(2)
1.0xc3x97P less than xcex4 less than 5.0xc3x97Pxe2x80x83xe2x80x83(3)
nL/nT less than 0.25xe2x80x83xe2x80x83(4)
where P is the pixel pitch (given in micrometers) of the image pickup device; N is the number of pixels (given in million) of the image pickup device; xcex4 is the maximum value (given in micrometers) of the absolute value of lateral chromatic aberration for a wavelength of 620 nanometers or 460 nanometers in an area containing the center of the image plane and accounting for 90% of the whole image plane; and nT nL number of lens elements constituting the optical system and the number of lens elements made of anomalous dispersion glass, respectively.
In this case, it is desirable for the optical system to have a detachable structure.
It is preferable to satisfy the following conditions:
1.5 less than P less than 4.0xe2x80x83xe2x80x83(5)
4.0 less than N less than 20xe2x80x83xe2x80x83(6)
In that case, it is even more desirable to satisfy the following condition:
6.0 less than N less than 20xe2x80x83xe2x80x83(7)
It is preferable that the image pickup device should be a device having a matrix or mosaic color filter array provided over a photoelectric conversion surface and should satisfy the following condition:
0 less than 1.3xc3x97P+4xe2x80x83xe2x80x83(8)
where xcex8 is the angle at which the central ray of a light beam exiting from the optical system or a ray passing through the center of a stop is incident on the image pickup device at each point within 90% of the image circle.
In this case, it is even more desirable to satisfy the following condition:
xcex8 less than P+4xe2x80x83xe2x80x83(9)
In addition, it is preferable to satisfy the following condition:
1.5xc3x97P less than 6 less than 5.0xc3x97Pxe2x80x83xe2x80x83(10)
In addition, it is preferable to satisfy the following condition:
nL/nT less than 0.20xe2x80x83xe2x80x83(11)
The reasons for adopting the above-described arrangement in the present invention, together with the function thereof, will be described below.
Assuming that the standard angular resolution of the human eye is 1xe2x80x2 (corresponding to the visual acuity 1.0), because this is a value that provides the critical resolution, resolution at which a line-and-space pattern contained in an image can be recognized as a sharp image is considered to be lower (larger in angle) than the above-described angular resolution.
For example, if the solid angle of the resolution is assumed to be about 1.67xe2x80x2 (corresponding to the visual acuity 0.6), when an image is observed from a position 40 centimeters away from it, for example, it can be recognized as a sharp image down to a size of 0.194 millimeters.
When the image can be recognized as a sharp image as stated above, the above-described color shift can be clearly recognized if it occurs. Accordingly, to make the color shift inconspicuous in this case, the lateral chromatic aberration must be corrected to xc2xd or less of the pixel pitch.
When a print of A4 size (210 mmxc3x97297 mm) is observed from a position 40 centimeters away from it, the number of pixels required in the above-described case is about 1,100xc3x971,500=1,650,000 pixels. Considering that a margin of about 10 millimeters is left at each of the peripheral edges of ,the print, the number of pixels required is about 1,400,000 pixels.
Basically, numerical values hereinafter presented are determined on the assumption that a print of A4 size is observed from a position 40 centimeters away from it. When there is a linear relationship between the print size and the viewing distance, the number of pixels obtained does not depend on the print size. In general, however, as the print size decreases, the viewing distance tends to increase in comparison to the print size. Therefore, in the case of a small print size, e.g. A6 size (105 mmxc3x97149 mm), a smaller number of pixels than the presented number of pixels will suffice. In other words, in the present invention, an image that can be satisfactorily enlarged to a print size of the order of 8xc3x9710xe2x80x3 size or larger is regarded as a high-quality image comparable to silver halide photographs.
Incidentally, in the case of an image produced by using an image pickup device, pixel units that constitute the image are fixed naturally. Therefore, it is impossible to express an object having a spatial frequency higher than the pixel units. When a test chart having pattern portions with various spatial frequencies is photographed, patterns whose narrowest line width is equal to the pixel pitch are all observed entirely resolved. However, in the case of patterns having a spatial frequency higher than the pixel units, in which the line width is less than the pixel pitch and the narrowest line width is, for example, xc2xd of the pixel pitch, only low-frequency light and dark patterns, i.e. moire, can be observed.
In the case of an image with the number of pixels obtained by the above-described calculation, the smallest pixel units of the print can be clearly recognized by man with high contrast. Accordingly, when the results of the above-described test chart observation are extended to apply to an ordinary image, the image appears sharp up to a certain spatial frequency, but it rapidly becomes impossible to resolve image components having spatial frequencies higher than the smallest pixel units. Therefore, although the contrast of the image is high, the image is poor in gradation representation and appears rough.
To obtain an image of abundant gradation, it is necessary to represent spatial frequencies up to one close to the resolution limit for the human perception.
In this regard, we have already disclosed that it is necessary in order to obtain image quality equal to that of silver halide photographs to use pixel units corresponding to a solid angle of at least 1.25xe2x80x2 (visual acuity 0.8), and about 2,500,000 pixels are needed when a margin of about 10 millimeters is left for A4 size.
The foregoing is a discussion taking notice of resolution. As has been stated above, the problem concerning color shift also arises from the fundamental problem that the displaced portion becomes stairstepped owing to the pixel units. Therefore, when the light intensity is gradient, if the gradient can be represented even more faithfully instead of being averaged, the color shift is not a problem of the image pickup device but a problem associated solely with the lens system, which is totally equivalent to that in the case of silver halide photographs.
To set a level for faithfully representing the gradient of the light intensity, it is possible to consult the above-described evaluation of the human sense of sight with respect to resolution. That is, when a print is observed, it will appear the same to human beings if the light intensity is represented in excess of pixel units that human beings cannot recognize at all. Conversely, if the observer can clearly recognize pixel units, it is impossible to recognize that changes in light intensity are represented faithfully, as stated above.
It was found that in the case of resolution, a solid angle of at least 1.25xe2x80x2 (visual acuity 0.8) is needed, as stated above, whereas, in the case of color shift, a larger number of pixels than the number of pixels obtained in the examination of resolution is needed because there are not only changes in intensity but also changes in hue. More specifically, it was found that if an image is composed of pixel units smaller than a solid angle of about 1.11xe2x80x2 (visual acuity 0.9), even if color shift caused by lateral chromatic aberration of a lens is represented in a stairstep shape by the pixel units, human beings cannot practically perceive the difference between the stairstep changes in intensity and the changes in intensity in a case where the color shift caused by the lens are represented faithfully.
When the solid angle of 1.11xe2x80x2 is converted into the corresponding number of pixels by a calculation similar to the above, it is about 3 million pixels when a margin of 10 millimeters is left at each of the peripheral edges of a print of A4 size.
We examined the allowable limit of the lateral chromatic aberration of the lens in the above-described case, and found that even if the lateral chromatic aberration is greater than the pixel units, human beings cannot clearly recognize color shift because it is generally accompanied by changes in intensity.
In other words, if an image pickup device having 3 million pixels or a larger number of pixels is used, even if there is a lateral chromatic aberration 1.0 or more times larger than the pixel pitch, human beings cannot clearly recognize the color shift.
However, if the lateral chromatic aberration of the lens is 5 or more times larger than the pixel units, the amount of shift caused by the chromatic aberration becomes excessively large, and it becomes impossible to permit the image quality deterioration caused by the color shift even if the effect of photographing with the image pickup device is ignored.
These conditions are expressed by the lower limit of the following condition (2) and condition (3).
3.0 less than N less than 20xe2x80x83xe2x80x83(2)
1.0xc3x97P less than xcex4 less than 5.0xc3x97Pxe2x80x83xe2x80x83(3)
where P is the pixel pitch (given in micrometers) of the image pickup device; N is the number of pixels (given in million) of the image pickup device; and xcex4 is the maximum value (given in micrometers) of the absolute value of lateral chromatic aberration for a wavelength of 620 nanometers or 460 nanometers in an area containing the center of the image plane and accounting for 90% of the whole image plane.
The reason why the condition concerning the lateral chromatic aberration is limited within an area containing the center of the image plane and accounting for 90% of the whole image plane is that the remaining area consists substantially of the four corners of the image plane, and it may be supposed that even if color shift occurs in the remaining area, it is unlikely to present a problem in terms of deterioration of the image quality. The reason why the wavelength for which lateral chromatic aberration is obtained is set at 620 nanometers or 460 nanometers is as follows. The e-line, which is handled as a reference welength, is considered to be a representative of a green image, and the above-described wavelengths are selected as representative wavelengths in view of the spectral characteristics of red and blue. The representative wavelengths shift according to the spectral transmittance of color filters used in the apparatus and the spectral sensitivity of the image pickup device. However, there will be no problem in terms of optical design if the representative wavelengths are regarded as being on the order of the above-mentioned wavelengths.
The upper limit of the condition (2), i.e. 20, is determined by taking into consideration the fact that as the number of pixels increases, the image reading speed and the recording speed to a recording medium become excessively low. The upper limit value of the condition (2) corresponds to the above-described solid angle of 0.0513xe2x80x2 (visual acuity 2.3), which is determined by taking into consideration the fact that at the upper limit value, even if the image is enlarged to A3 or larger, for example, and thus the viewing distance becomes short in comparison to the print size, the observer cannot practically recognize deterioration of the image quality.
To produce an image pickup device having the above-described number of pixels, it is desirable to satisfy the following condition (1):
1.5 less than P less than 7.0xe2x80x83xe2x80x83(1)
The upper limit of the condition (1), i.e. 7.0, is set by taking into consideration the cost of the image pickup apparatus. If P is not smaller than the upper limit value, the size of the image pickup device becomes large in comparison to the number of pixels. Consequently, the number of devices obtained from one wafer decreases, causing an increase in cost. The lower limit value of the condition (1), i.e. 1.5, is set by taking into consideration both the image quality deterioration due to the diffraction limit of the optical system and the production cost of the optical system. If P is not larger than the lower limit value, it becomes impossible to ignore the effect of diffractive blurring in a case where the number of pixels is of the order of 3 million even if the optical system has a diffraction limit level of performance. This causes the image quality to be degraded. In the case of using an image pickup device in which the number of pixels exceeds 10 million, if the optical system has performance close to the diffraction limit, the image quality required for a print can be satisfactorily obtained. However, the manufacturing tolerances of each element constituting the optical system become tightened, causing an increase in the processing cost unfavorably.
Adopting a combination of the above-described image pickup device and optical system provides an allowance for the correction of lateral chromatic aberration when the optical system is constructed. Consequently, it is possible to reduce the number of lens elements constituting the optical system and to minimize the number of lens elements using anomalous dispersion glass. Accordingly, it becomes possible to construct a compact and low-cost optical system.
More specifically, the number of lens elements using anomalous dispersion glass can be reduced to 25% or less of the total number of lens elements constituting the optical system. Thus, it becomes possible to reduce the processing cost of the optical system. The condition for this matter is given by the following expression (4):
nL/nT less than 0.25xe2x80x83xe2x80x83(4)
where nT and nL are the number of lens elements constituting the optical system and the number of lens elements made of anomalous dispersion glass, respectively.
The image pickup apparatus according to the present invention may be arranged such that color filters of different colors are successively set on the object side of the apparatus or in the apparatus, and the photographed images are superimposed on one another to construct a color image. Alternatively, two or more image pickup devices are used, and optical paths are led to the respective image pickup devices by an optical path splitting prism, for example, thereby simultaneously photographing images of a plurality of different colors.
The image pickup apparatus may also be arranged such that a device having a matrix or mosaic color filter array provided over a photoelectric conversion surface is used to photograph an image, and the photographed image is subjected to a matrix computation to construct a color image.
The latter system is inferior in image quality to the former system because it needs interpolation of the image by a matrix computation. However, the latter system uses a smaller number of image pickup devices and is capable of obtaining a color image by a smaller number of photographing operations. Therefore, the latter system is superior in terms of cost and ease of use. Accordingly, this system is adopted in many products, particularly consumer products.
In this system, each pixel of the image pickup device comprises a combination of a photoelectric conversion surface region and a color filter element provided thereover. Therefore, if light passing through a color filter element enters a photoelectric conversion surface region different from one at which the light should arrive, information concerning color cannot correctly be obtained. Therefore, it is desirable for light passing through the optical system to be perpendicularly incident on the image pickup device. However, if a light-blocking portion or the like is provided around each photoelectric conversion surface, even if the light is incident on the image pickup device at a certain angle, the effect on the image can be ignored. However, if the light-blocking portion is widened extremely, the area of the photoelectric conversion surface with respect to the whole area decreases. Consequently, the amount of light that can be converted into an electric signal reduces, causing reduction in sensitivity.
The following condition (8) is a condition concerning the allowable range of the angle of incidence on the image pickup device in which the effect on the image can be ignored when a necessary opening ratio is assumed from the viewpoint of sensitivity.
0 less than 1.3xc3x97P+4xe2x80x83xe2x80x83(8)
where xcex8 is the angle at which the central ray of a light beam exiting from the optical system or a ray passing through the center of the stop is incident on the image pickup device at each point within 90% of the image circle.
If xcex8 is not smaller than the upper limit of the condition (8), i.e. 1.3xc3x97P+4, the error of color information obtained at the peripheral portion of the image plane becomes excessively large in comparison to the center of the image plane where the light beam is approximately perpendicularly incident on the image pickup device for the reason stated above, giving rise to a problem that the color in the center of the image plane and the color in the peripheral portion do not agree with each other even if electrical processing is used.
To obtain an image of higher quality, it is preferable to further reduce the pixel units of the image. In this case, it is preferable to satisfy the following condition:
4.0 less than N less than 20xe2x80x83xe2x80x83(6)
It is even more desirable to satisfy the following condition:
6.0 less than N less than 20xe2x80x83xe2x80x83(7)
If the condition (7) is satisfied, an image of even higher quality can be obtained.
Regarding the pixel pitch, it is preferable to satisfy the following condition:
1.5 less than P less than 4.0xe2x80x83xe2x80x83(5)
An effective way of constructing the optical system at further reduced cost is to loosen the restriction on lateral chromatic aberration. It is preferable to satisfy the following condition:
1.5xc3x97P less than 5 less than 5.0xc3x97Pxe2x80x83xe2x80x83(10)
It is preferable from the viewpoint of further reducing the cost that the number of lens elements using anomalous-dispersion glass among the lens elements constituting the optical system should satisfy the following condition:
nL/nT less than 0.20xe2x80x83xe2x80x83(11)
It is even more desirable that the angle at which a ray exiting from the optical system is incident on the image pickup device should satisfy the following condition:
xcex8 less than P+4xe2x80x83xe2x80x83(9)
In the image pickup apparatus according to the present invention, the apparatus body and the optical system may be integrated in one unit. Alternatively, the optical system is arranged to be replaceable.
Still other objects and advantages of the invention will in part be obvious and will in part be apparent from the specification.
The invention accordingly comprises the features of construction, combinations of elements, and arrangement of parts which will be exemplified in the construction hereinafter set forth, and the scope of the invention will be indicated in the claims.